CN110795465A

CN110795465A - User scale pre-estimation method, device, server and storage medium

Info

Publication number: CN110795465A
Application number: CN201910871680.7A
Authority: CN
Inventors: 成昊; 谢思发; 程序; 张涵宇; 刘文强; 江小琴
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2019-09-16
Filing date: 2019-09-16
Publication date: 2020-02-14
Anticipated expiration: 2039-09-16
Also published as: CN110795465B

Abstract

The invention provides a user scale pre-estimation method, a user scale pre-estimation device, a server and a storage medium, and belongs to the technical field of computers. The method comprises the following steps: for each application label of the first target application program, acquiring a plurality of label data corresponding to the application label, and generating a first application characteristic vector of the first target application program according to the plurality of label data; acquiring second application feature vectors of a plurality of online second target application programs and first user weights of a plurality of first users for each second target application program respectively; determining a first user feature vector of each first user according to the second application feature vector and the first user weight; and inputting the first application characteristic vector and the first user characteristic vector into a scale pre-estimation model to obtain the scale of the first user. Due to the introduction of the first application characteristic vector and the first user characteristic vector, the attribute of the first target application program can be comprehensively described, the preference of the user is reflected, and the estimated first user scale is more accurate.

Description

User scale pre-estimation method, device, server and storage medium

Technical Field

The present invention relates to the field of computer technologies, and in particular, to a method and an apparatus for estimating a user scale, a server, and a storage medium.

Background

In order to provide better user experience for users, for a new game at the initial stage of coming-on or coming-on, the user scale of the new game needs to be estimated, and resources needing to be invested in the new game are determined according to the estimation result of the user scale.

In the related technology, before the user scale of a new game is estimated, user scale data in the initial stage of online game in a game large disc and user scale data in the appointed time after the game is online need to be acquired, and a depth time sequence model is obtained through training according to the user scale data in the initial stage of online game and the user scale data in the appointed time after the game is online. When the user scale of the new game is estimated, user scale data of the new game in the online initial stage, which needs to be estimated, are obtained, the user scale data of the new game in the online initial stage are input into the depth time sequence model, and the user scale data of the new game in the appointed time after the new game is online are obtained.

In the related technology, the estimation is only carried out according to the scale data of the user, and the estimation result is not accurate enough.

Disclosure of Invention

The embodiment of the invention provides a user scale pre-estimation method, a user scale pre-estimation device, a server and a storage medium, which can solve the problem of low user scale pre-estimation accuracy. The technical scheme is as follows:

according to an aspect of the embodiments of the present invention, there is provided a user scale prediction method, including:

acquiring a plurality of application tags of a first target application program, wherein each application tag corresponds to a plurality of options;

for each application label, acquiring a plurality of label data corresponding to the application label, wherein each label data comprises a selected target option in a plurality of options corresponding to the application label;

generating a first application characteristic vector of the first target application program according to the plurality of application labels and label data corresponding to each application label;

acquiring second application feature vectors of a plurality of online second target application programs and first user weights of a plurality of first users for each second target application program respectively;

determining a first user feature vector of each first user according to the second application feature vector of each second target application program and the first user weight of each first user for each second target application program;

and inputting the first application characteristic vector and the first user characteristic vector of each first user into a scale pre-estimation model to obtain the first user scale of the first target application program.

In one possible implementation manner, the inputting the first application feature vector and the first user feature vector of each first user into a scale prediction model to obtain a first user scale of the first target application includes:

acquiring attribute information of each first user;

determining a second user feature vector of each first user according to the attribute information of each first user;

and inputting the first application characteristic vector, the first user characteristic vector of each first user and the second user characteristic vector into the scale pre-estimation model to obtain the first user scale of the first target application program.

In another possible implementation manner, each tag data further includes a user identifier of a second user who selects the target option;

generating a first application feature vector of the first target application program according to the plurality of application tags and the tag data corresponding to each application tag, including:

for each second user, determining a target option selected by the second user from multiple options corresponding to multiple application labels according to the user identification of the second user, and obtaining multiple target options;

generating a third application feature vector corresponding to the second user according to the plurality of target options;

and carrying out weighted summation on the third application characteristic vector corresponding to each second user to obtain the first application characteristic vector.

In another possible implementation manner, before the obtaining of the plurality of tag data corresponding to the application tag, the method further includes:

for each application label, establishing a questionnaire system according to the application label;

receiving a numerical value of the application label scored on the questionnaire system by a second user;

according to the numerical value, selecting a target option corresponding to the numerical value from a plurality of options corresponding to the application label;

forming label data by the application identifier of the first target application program, the application label and the target option, and storing the label data in a label database;

the obtaining of the plurality of tag data corresponding to the application tag includes:

and acquiring a plurality of label data corresponding to the application label from the label database according to the application identifier of the first target application program and the application label.

In another possible implementation manner, after the inputting the first application feature vector and the first user feature vector of each first user into the scale prediction model to obtain the first user scale of the first target application, the method further includes:

acquiring statistical data of the first target application program;

determining a statistical feature vector of the first target application program according to the statistical data;

and inputting the statistical feature vector and the first user scale into a later-stage correction model to obtain a second user scale of the first target application program.

In another possible implementation manner, before the inputting the first application feature vector and the first user feature vector of each first user into the scale prediction model to obtain the first user scale of the first target application, the method further includes:

acquiring fourth application characteristic vectors of a plurality of online third target application programs;

for each third target application program, acquiring a third user feature vector and a learning label of a plurality of third users for the third target application program, wherein the learning label of each third user for the third target application program is used for indicating whether a terminal of the third user runs the third target application program or not;

and performing model training according to the fourth application characteristic vector of each third target application program, the third user characteristic vector of each third user and the learning label of each third user to obtain the scale pre-estimation model.

In another possible implementation manner, the performing model training according to the fourth application feature vector of each third target application, the third user feature vector of each third user, and the learning label of each third user to obtain the scale prediction model includes:

acquiring attribute information of each third user;

determining a fourth user feature vector of each third user according to the attribute information of each third user;

and performing model training according to the fourth application characteristic vector of each third target application program, the third user characteristic vector of each third user, the fourth user characteristic vector and the learning label of each third user to obtain the scale pre-estimation model.

According to another aspect of the embodiments of the present invention, there is provided a user scale prediction apparatus, including:

the acquisition module is used for acquiring a plurality of application tags of the first target application program, and each application tag corresponds to a plurality of options; for each application label, acquiring a plurality of label data corresponding to the application label, wherein each label data comprises a selected target option in a plurality of options corresponding to the application label;

a generating module, configured to generate a first application feature vector of the first target application program according to the multiple application tags and tag data corresponding to each application tag;

the obtaining module is further configured to obtain second application feature vectors of the plurality of second target application programs that have been online, and first user weights of the plurality of first users for each second target application program;

a determining module, configured to determine a first user feature vector of each first user according to the second application feature vector of each second target application and a first user weight of each first user for each second target application;

and the pre-estimation module is used for inputting the first application characteristic vector and the first user characteristic vector of each first user into a scale pre-estimation model to obtain the first user scale of the first target application program.

In a possible implementation manner, the pre-estimation module is further configured to obtain attribute information of each first user; determining a second user feature vector of each first user according to the attribute information of each first user; and inputting the first application characteristic vector, the first user characteristic vector of each first user and the second user characteristic vector into the scale pre-estimation model to obtain the first user scale of the first target application program.

In another possible implementation manner, each tag data further includes a user identifier of a second user who selects the target option; the generating module is further configured to determine, for each second user, a target option selected by the second user from the multiple options corresponding to the multiple application tags according to the user identifier of the second user, so as to obtain multiple target options; generating a third application feature vector corresponding to the second user according to the plurality of target options; and carrying out weighted summation on the third application characteristic vector corresponding to each second user to obtain the first application characteristic vector.

In another possible implementation manner, the establishing module is configured to establish, for each application tag, a questionnaire system according to the application tag;

the receiving module is used for receiving a numerical value of the application label scored by the second user on the questionnaire system;

the selecting module is used for selecting a target option corresponding to the numerical value from a plurality of options corresponding to the application label according to the numerical value;

the storage module is used for forming label data by the application identifier of the first target application program, the application label and the target option and storing the label data into a label database;

the obtaining module is further configured to obtain, from the tag database, a plurality of tag data corresponding to the application tag according to the application identifier of the first target application program and the application tag.

In another possible implementation manner, the obtaining module is further configured to obtain statistical data of the first target application;

the determining module is further configured to determine a statistical feature vector of the first target application according to the statistical data;

the pre-estimation module is further configured to input the statistical feature vector and the first user scale into a later-stage correction model to obtain a second user scale of the first target application program.

In another possible implementation manner, the obtaining module is further configured to obtain fourth application feature vectors of a plurality of third target application programs that have been online; for each third target application program, acquiring a third user feature vector and a learning label of a plurality of third users for the third target application program, wherein the learning label of each third user for the third target application program is used for indicating whether a terminal of the third user runs the third target application program or not;

and the model training module is used for carrying out model training according to the fourth application characteristic vector of each third target application program, the third user characteristic vector of each third user and the learning label of each third user to obtain the scale pre-estimation model.

In another possible implementation manner, the model training module is further configured to obtain attribute information of each third user; determining a fourth user feature vector of each third user according to the attribute information of each third user; and performing model training according to the fourth application characteristic vector of each third target application program, the third user characteristic vector of each third user, the fourth user characteristic vector and the learning label of each third user to obtain the scale pre-estimation model.

According to another aspect of the embodiments of the present invention, there is provided a server, including: a processor and a memory, wherein the memory stores at least one instruction, at least one program, a set of codes, or a set of instructions, which are loaded and executed by the processor to implement the operations of the user scale prediction method according to any one of the above possible implementations.

According to another aspect of embodiments of the present invention, there is provided a computer-readable storage medium having stored therein at least one instruction, at least one program, set of codes, or set of instructions, which is loaded by a processor and has an operation to implement as performed in the user size estimation method.

In the embodiment of the invention, a plurality of application tags of a first target application program are obtained, and each application tag corresponds to a plurality of options; for each application label, acquiring a plurality of label data corresponding to the application label, wherein each label data comprises a selected target option in a plurality of options corresponding to the application label; generating a first application characteristic vector of a first target application program according to the plurality of application labels and label data corresponding to each application label; acquiring second application feature vectors of a plurality of online second target application programs and first user weights of a plurality of first users for each second target application program respectively; determining a first user feature vector of each first user according to the second application feature vector of each second target application program and the first user weight of each first user for each second target application program; and inputting the first application characteristic vector and the first user characteristic vector of each first user into a scale pre-estimation model to obtain the first user scale of the first target application program. In the embodiment of the invention, the first application characteristic vector and the first user characteristic vector are introduced, so that the attribute of the first target application program can be more comprehensively described, and the preference of the user can be embodied, thereby enabling the scale of the first user obtained through the scale pre-estimation model to be more accurate.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

FIG. 1 is a schematic illustration of an implementation environment provided by an embodiment of the invention;

fig. 2 is a flowchart of a method for estimating a user scale according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of an option for applying a tag provided by an embodiment of the invention;

FIG. 4 is a diagram illustrating database schema partitioning according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of a questionnaire system provided by an embodiment of the invention;

FIG. 6 is a diagram illustrating conversion of tag data into application feature vectors according to an embodiment of the present invention;

FIG. 7 is a diagram of a target application tag according to an embodiment of the present invention;

FIG. 8 is a schematic diagram of a user associated with a target application according to an embodiment of the present invention;

fig. 9 is a schematic structural diagram of a user scale estimation apparatus according to an embodiment of the present invention;

fig. 10 is a schematic structural diagram of a terminal according to an embodiment of the present invention;

fig. 11 is a schematic structural diagram of a server according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

With the development of computer technology, application programs are applied in the aspects of people's life, and the user scale of one application program often represents the value of the application program. Therefore, it is of great significance to pre-estimate the user scale of the application program and determine the resources to be invested in the application program according to the user scale of the application program. The application that needs to be scaled may be a gaming application, a social application, a video application, etc.

The embodiment of the invention is applied to a scene of user scale estimation of game application, takes a target application program as the game application as an example, and trains according to the specific game features of the game application and the features of game users to obtain a scale estimation model. When the scale of the game application needs to be pre-estimated, acquiring the game duration of other games of the user in the game board within a specified time period close to the online time according to the online time of the game application needing to be pre-estimated, and determining the characteristics of other game users in the game board according to the game duration of other games of the user in the game board and the game characteristics of the corresponding games; inputting the specific game characteristics of the game application needing scale pre-estimation and the characteristics of other game users in the game large disc into a scale pre-estimation model to obtain the user scale of the game application needing scale pre-estimation. Therefore, according to the obtained user scale, propaganda resources, promotion resources and the like needing to be invested in the game application can be determined, and according to the user characteristics in the obtained user scale, propaganda and promotion channels and the like which are effective to the game can be determined.

FIG. 1 is a schematic diagram of an implementation environment provided by an embodiment of the invention. Referring to fig. 1, the implementation environment includes a terminal 101 and a server 102.

The terminal 101 may be a computer, a mobile phone, a tablet computer or other electronic devices. The server 102 may be a server, a server cluster composed of several servers, or a cloud computing service center.

The terminal 101 and the server 102 are connected via a wireless or wired network. Moreover, a client that the server 102 provides services may be installed on the terminal 101, and a user corresponding to the terminal 101 may implement functions such as data transmission and message interaction through the client. The client may be a client installed on the terminal 101 that includes a questionnaire system. For example, the client may be a questionnaire application, a game application, or a browser, etc.

The questionnaire system comprises game application identification, application tags and options, wherein each application tag corresponds to a plurality of options. For example, the application tag is "game difficulty selection", and the options corresponding to the application tag may be "simple", "general", and "difficult". When the questionnaire system is run on the terminal 101, the user can select the target option corresponding to the application tag according to the characteristics of the game application. The user may be a game plan or a game developer, or may be a general user, that is, a general player of the game.

The server 102 may receive the tag data sent by the questionnaire system, and determine an application feature vector of the game application corresponding to the game application identifier according to the tag data. Determining a user characteristic vector according to the application characteristic vector of the game application and the game duration of the user; determining a scale pre-estimation model according to the application characteristic vector of the game application and the corresponding user characteristic vector; and obtaining the estimated user scale of the game application according to the application characteristic vector of the game application needing scale estimation, the corresponding user characteristic vector and the scale estimation model.

Fig. 2 is a flowchart of a method for estimating a user scale according to an embodiment of the present invention. Referring to fig. 2, the user scale prediction method includes the following steps:

201. the server obtains a plurality of application tags of the first target application program, wherein each application tag corresponds to a plurality of options.

The first target application program may be a game application that is about to be online or needs to be pre-scaled at the beginning of online.

In one possible implementation, the server obtains a plurality of application tags of the first target application, where the application tags embody characteristics of the game application, for example, referring to fig. 3, the application tags are "primary item class", "defective item class", "drawing", and the like. Each application tag corresponds to a plurality of options, for example, when the application tag is a "main item class", the plurality of options corresponding to the application tag may be "Action", "role playing RPG", "Competitive", or the like; when the application tag is a "secondary type", the multiple options corresponding to the application tag may be "general Action", "Action Adventure Action advance", "street Arcade", and the like; when the application tag is "draw a wind", a plurality of options corresponding to the application tag may be "realistic", "cartoon", "lovely", and the like.

In another possible implementation manner, before the server obtains the plurality of application tags of the first target application, the application tag of the first target application may be further set by the following steps: constructing a game knowledge graph, wherein the game knowledge graph comprises a plurality of game applications; the database schema for building the game knowledge graph, for example, the database schema includes 23 subsystems, see fig. 4, which includes 23 subsystems such as "PVP system", "combat system", "social system", "play system", and the like. Each subsystem includes a plurality of application tags, for example, the "play system" includes "main item type", "defective item type", "play time consumption", "attention consumption degree", "hand difficulty", "game content number", and the like. The server may obtain a plurality of application tags for the first target application based on the database schema of the game knowledge graph.

202. For each application label, the server obtains a plurality of label data corresponding to the application label, and each label data comprises a selected target option in a plurality of options corresponding to the application label.

In a possible implementation manner, the server may directly obtain tag data corresponding to the application tag, where each tag data includes a selected target option from a plurality of options corresponding to the application tag.

In another possible implementation manner, before the server obtains the plurality of tag data corresponding to the application tag, the following steps may be further performed: for each application label, the server establishes a questionnaire system according to the application label; receiving a numerical value of the second user scoring the application label on the questionnaire system; according to the numerical value, selecting a target option corresponding to the numerical value from a plurality of options corresponding to the application label; and forming label data by the application identifier, the application label and the target option of the first target application program, and storing the label data in a label database.

The server establishes a questionnaire system according to the application tags, wherein the questionnaire system comprises the application tags and a plurality of numerical values corresponding to each application tag, and different numerical values are used for identifying different options.

An example of selecting an application tag and a numerical value of the questionnaire system is shown in fig. 5, where the application tag may be "game difficulty selection", "whether the game view angle can be adjusted", and the like, and different numerical values identify different options, for example, when the application tag is "game difficulty selection", a numerical value "1" may represent "simple", a numerical value "2" may represent "general", and a numerical value "3" may represent difficulty; when the application tag is "whether or not the game view is adjustable", a value "1" may indicate "yes", and a value "0" may indicate "no".

The second user may score each application tag based on the characteristics of the application tag. The second user can select the corresponding numerical value of the application label in the questionnaire system according to the characteristics of the application label; the second user may also input a numerical value corresponding to each application tag in the questionnaire system according to the characteristics of the application tag. The server selects a target option corresponding to the numerical value from a plurality of options corresponding to the application label according to the numerical value returned by the questionnaire system; and forming label data by the application identifier, the application label and the target option of the first target application program, and storing the label data in a label database. Correspondingly, the step of acquiring, by the server, a plurality of tag data corresponding to the application tag may further be: and acquiring a plurality of label data corresponding to the application label from a label database according to the application identifier and the application label of the first target application program.

In another possible implementation manner, the target option corresponding to each application tag is set according to the characteristics of the first target application program. For example, the application tag is a fact type application tag such as "production company", "release time", and the like, and the server can acquire a target option of the fact type application tag through the crawler system.

In the embodiment of the invention, the server directly determines the target option of the fact application label through the crawler system, so that the human participation can be reduced, and the complexity of acquiring the target option of the application label is reduced.

203. The server generates a first application feature vector of the first target application program according to the plurality of application tags and tag data corresponding to each application tag.

In one possible implementation manner, the first target application includes a plurality of application tags for describing characteristics of the first target application, each application tag corresponds to a target option, and each application tag and the target option of the application tag form one tag data, that is, the first target application corresponds to a plurality of tag data. Accordingly, the step of generating, by the server, the first application feature vector of the first target application may be: for each application label of the first target application program, determining the number of a plurality of options corresponding to the application label, and taking the number of the options corresponding to the application label as the dimension of a first vector corresponding to the application label; the position 1 and other positions 0 of the target option in the first vector are represented; splicing the first vectors corresponding to each application label to generate a second vector; and for a plurality of second vectors corresponding to the first target application program, summing each bit of the plurality of second vectors, and averaging to obtain the first application characteristic vector of the first target application program.

For example, if the application tag is "number of players", the application tag corresponds to "single player", "multi-player single player", "single online", "multi-player online", and "other" 5 options, and the target option corresponding to the application tag is "multi-player single player", the first vector corresponding to the application tag is a 5-dimensional vector, which can be represented as "0 |1|0|0|0 |.

In another possible implementation manner, each tag data includes, in addition to a selected target option in the multiple options corresponding to the application tag, a user identifier of a second user who selects the target option; correspondingly, the step of generating, by the server, the first application feature vector of the first target application program according to the plurality of application tags and the tag data corresponding to each application tag may further be: for each second user, the server determines a target option selected by the second user from the multiple options corresponding to the multiple application labels according to the user identification of the second user, and obtains multiple target options; generating a third application characteristic vector corresponding to the second user according to the plurality of target options; and carrying out weighted summation on the third application characteristic vectors corresponding to each second user and then averaging to obtain the first application characteristic vectors.

The server may obtain tag data from a tag database, each tag data including an application identification, a user identification, an application tag, and a target option. The application identifier is an application identifier of a first target application program, the application tag is an application tag of the first target application program, the target option is a selected target option in a plurality of options corresponding to each application tag, and the user identifier is a user identifier of a second user who selects the target option. The second user can select a target option matched with the application tag of the first target application program from a plurality of options corresponding to the application tag of the first target application program according to the characteristics of the first target application program in the questionnaire system, the server can determine a target option selected by the second user from a plurality of options corresponding to each application tag according to the user identification of the second user, and the application identification of the first target application program, the user identification of the second user, the application tag and the target option selected by the second user form tag data; for each application label of the first target application program, determining the number of a plurality of options corresponding to the application label, and taking the number of the options corresponding to the application label as the dimension of a first vector corresponding to the application label; the position 1 and other positions 0 of the target option in the first vector are represented; for each second user, the first vectors corresponding to each application label are spliced to generate a third application feature vector corresponding to each second user; determining the weight of each second user according to the user identification of the second user, and multiplying the weight of each second user by each bit of the third application characteristic vector corresponding to each second user according to the weight of each second user and the third application characteristic vector corresponding to each second user to obtain a third vector; and adding and summing the plurality of third vectors, and then averaging to obtain a first application characteristic vector of the first target application program.

The second user may be a planning or developing person of the first target application program, or may also be a general user, and the planning or developing person of the first target application program may have a higher weight, and the general user may have a lower weight.

Referring to fig. 6, the server may obtain tag data from a tag database and convert the tag data into a first vector. For example, an application identifier "Ga" and a user identifier "Us" in the tag data are marked with a "clear font", a target option is "yes", correspondingly, the tag data are marked with a "Ga, Us" and a clear font, yes ", a first vector corresponding to the tag data is" clear font: 0|1 "; the application identifier "Ga" and the user identifier "Ud" in the tag data, the application tag is "font clear", the target option is "yes", correspondingly, the tag data is "Ga, Ud, the font clear, yes", the first vector corresponding to the tag data is "font clear: 0|1 "; the application identification "Ga" and the user identification in the tag data are "Us", the application tag is "main item class", the target option is "shooting | survival", correspondingly, the tag data are "Ga, Us, main item class, shooting | survival", a first vector corresponding to the tag data is "main item class: 0|0|1 … 0|1 "; the application identifier "Ga" and the user identifier "Ud" in the tag data, the application tag is "master category", the target option is "shooting | action", correspondingly, the tag data is "Ga, Ud, master category, shooting | action", a first vector corresponding to the tag data is "master category: 0|1|0 … 0|1 ".

And for each second user, the server splices the first vectors corresponding to each application label to obtain a third application feature vector corresponding to each second user. Accordingly, with continued reference to fig. 6, the third user feature vector corresponding to the second user "Us" is "0 |1 … 0|0|1 … 0| 1", and the third user feature vector corresponding to the second user "Ud" is "0 |1 … 0|1|0 … 0|0| 1".

The server multiplies the weight of each second user by each position of a third application characteristic vector corresponding to each second user to obtain a third vector; and adding and summing the plurality of third vectors, and then averaging to obtain a first application characteristic vector of the first target application program. With continued reference to FIG. 6, for example, the second user "Us" and the second user "Ud" are both weighted at 0.5, and accordingly, the first application feature vector of the first target application, identified as "Ga", is "0 |1 … 0|0.5|0.5 … 0| 1".

In the embodiment of the invention, the server acquires a plurality of label data corresponding to the application label according to the target options selected by a plurality of second users on the questionnaire system, different weights are given to different second users, and the first application characteristic vector is determined according to the third user characteristic vectors corresponding to different second users and the weight corresponding to each second user, so that the confidence coefficient of the label data can be enhanced.

204. The server acquires second application characteristic vectors of a plurality of second target application programs which are online; the server obtains a first user weight of each second target application program for the plurality of first users.

The first user is a potential user of the first target application, and the third target application may be an online target application or a game application in a game knowledge graph. The steps of the server obtaining the second application feature vectors of the plurality of second target application programs that are online are similar to steps 201 to 203, and are not described herein again.

In one possible implementation manner, the server may determine a plurality of first users according to the online time of the first target application, and accordingly, the step of determining the plurality of first users by the server may be: the server determines a specified time period close to the online time of the first target application program, and determines users of a second target application program which are online in the specified time period as a plurality of first users.

In another possible implementation manner, the server may further sample the obtained plurality of first users, and accordingly, the step of determining, by the server, the plurality of first users may be: the server determines a specified time period close to the online time of the first target application program, samples users of a second target application program which are online in the specified time period, and determines a plurality of first users.

After the server determines the plurality of first users, a first user weight for each second target application is further determined, and accordingly, the step of determining the first user weight for each second target application by the server may be: for each first user, the server obtains the running time length of the first user for each second target application program, and determines the first user weight of each first user for each second target application program according to the running time length of the first user for each second application program.

For example, the specified time period is 30 days before the first target application is online, the running duration of the first user for the second target application "Ga" acquired by the server is 3 hours, and the running duration of the first user for the second target application "Gb" is 7 hours; the server may determine that the first user weight of the first user for the second target application "Ga" is 0.3 and the first user weight of the first user for the second target application "Gb" is 0.7.

205. The server determines a first user feature vector for each first user based on the second application feature vector for each second target application and the first user weight for each first user for each second target application.

The server maps the second application feature vector to the first user to obtain the first user feature vector of the first user, for example, referring to fig. 7, the second target application program is a game application, the plurality of second target application programs are respectively "game a", "game B", "game C", "game a" corresponds to "tag 1", game B corresponds to "tag 2", and game C "corresponds to" tag 3 "; referring to fig. 8, the first users are "player 1", "player 2", and "player 3", respectively, wherein "player 1" has run "game a" and "game B" within a specified time period that is close to the online time of the first target application, then "player 1" corresponds to "tag 1" and "tag 2"; "Player 2" has run through "Games B" and "Games C" within a specified time period that is close to the first target application online time, then "Player 2" corresponds to "Tab 2" and "Tab 3"; "Player 3" has run both "Game A" and "Game C" for a specified period of time that is close to the first target application online time, then "Player 3" corresponds to "Label 1" and "Label 3".

The server determines a first user feature vector for each first user based on the second application feature vector for each second target application and the first user weight for each first user for each second target application. For example, the step may be: the server determines a second application feature vector of a second target application program 'Ga' according to the step (1)

Second targetThe second application feature vector of the application program "Gb" is

If the server determines that the first user weight of the first user for the second target application "Ga" is 0.3 and the first user weight of the first user for the second target application "Gb" is 0.7 according to step (2), the server may determine the first user feature vector of each first user according to the following formula one:

the formula I is as follows:

wherein, U_aA first user is indicated and a second user is indicated,

a first user feature vector, G, representing a first user_aRepresenting a second application "Ga",

a second application feature vector representing a second target application "Ga"; g_bIndicating a second application program "Gb",

a second application feature vector representing a second target application program "Gb".

In the embodiment of the invention, the server determines the weight of each second target application program according to the running time of the first user for each second target application program in the specified time period which is close to the online time of the first target application program, and maps the first application characteristic vector to the first user characteristic vector, so that the application trend of the user to the target application program can be reflected more accurately in real time.

206. And the server inputs the first application characteristic vector and the first user characteristic vector of each first user into a scale pre-estimation model to obtain the first user scale of the first target application program.

When the scale of the first target application program needs to be pre-estimated, the server may input the first application feature vector and the first user feature vector of each first user into the scale pre-estimation model to obtain the first user scale of the first target application program. The scale pre-estimation model predicts whether the first user will run the first target application, wherein the scale of the first user is the number of the first users who will run the first target application.

In one possible implementation manner, the server splices the first application feature vector and the first user feature vector of each first user, and inputs the scale pre-estimation model to obtain the first user scale of the first target application program.

In another possible implementation manner, if the plurality of first users are obtained by sampling users of the second target application program which is actively online, the server determines that the ratio of the number of the users of the second target application program which is online to the number of the plurality of first users is a sampling rate; splicing the first application characteristic vector and the first user characteristic vector of each first user, and inputting a scale pre-estimation model to obtain the scale of a third user, wherein the scale of the third user is the number of the first users which can run the first target application program and are predicted by the scale pre-estimation model; and multiplying the third user scale by the sampling rate to obtain the first user scale of the first target application program.

The number of users in the game board can fluctuate, so that a certain error is allowed to exist in the sampling rate, the sampling rate can be represented as a numerical range according to a set error range, and correspondingly, the scale of the first user of the first target application program is also the corresponding numerical range.

In another possible implementation manner, the server splices the first application characteristic vector and the first user characteristic vector of each first user, and inputs a scale pre-estimation model to obtain the probability of each first user for operating the first target application program; for each first user, when the probability of the first user for running the first target application is greater than a specified threshold, determining that the first user will run the first target application, and when the probability of the first user for running the first target application is not greater than the specified threshold, determining that the first user will not run the first target application. The number of first users who will run the first target application is determined to be the first user size. The predetermined threshold may be a set threshold or a threshold obtained by a machine learning method.

In the embodiment of the invention, when the user scale is estimated, factors such as the characteristics of the first target application program, the preference of the user and the like are comprehensively considered, and the user scale estimation is converted into the problem of determining the conversion rate, so that the result of the user scale estimation is more accurate; and the user scale pre-estimation is carried out without acquiring the early-stage user scale data of the first target application program, the user scale pre-estimation does not depend on the issuing progress of the first target application program, and the timeliness is good.

In another possible implementation manner, the server may further obtain attribute information of each first user; determining a second user characteristic vector of each first user according to the attribute information of each first user; and inputting the first application characteristic vector, the first user characteristic vector of each first user and the second user characteristic vector into a scale pre-estimation model to obtain the first user scale of the first target application program.

The attribute information of the first user may be attribute information of the first user, such as gender, age, and the like. The server determines a second user characteristic vector of each first user according to the attribute information of each first user; and inputting the first application characteristic vector, the first user characteristic vector of each first user and the second user characteristic vector into a scale pre-estimation model to obtain the first user scale of the first target application program.

In the embodiment of the invention, the attribute information of the first user, such as gender, age and the like, is input into the scale estimation model for user scale estimation, so that the characteristics of the user are more comprehensively embodied, and the user scale can be more accurately estimated.

In another possible implementation manner, the server inputs the first application characteristic vector and the first user characteristic vector of each first user into a scale pre-estimation model, and after the first user scale of the first target application program is obtained, statistical data of the first target application program can be obtained; determining a statistical feature vector of the first target application program according to the statistical data; and inputting the statistical feature vector and the first user scale into a later-stage correction model to obtain a second user scale of the first target application program.

Wherein, the statistical data of the first target application program can be market research data, player questionnaire data, third party statistical data and the like; the later correction model is a machine learning model. The server abstracts the statistical data into corresponding statistical characteristic vectors, splices the statistical characteristic vectors with the first user scale obtained through the scale input model, and inputs the spliced statistical characteristic vectors into the later-stage correction model to obtain the second user scale of the first target application program.

In the embodiment of the invention, the traditional statistical data is effectively utilized, and the model is combined with the traditional data, so that the result of the scale estimation of the user is more accurate.

The server can directly obtain the trained scale pre-estimation model to pre-estimate the scale of the user; the scale pre-estimation model can also be obtained through model training in advance, and when the scale pre-estimation model is needed to carry out user scale pre-estimation, the scale pre-estimation model obtained through model training in advance is obtained to carry out user scale pre-estimation; before the step of user scale pre-estimation, a scale pre-estimation model can be obtained through model training, and then the user scale pre-estimation can be carried out according to the scale pre-estimation model. Correspondingly, the server obtains the scale pre-estimation model through model training, and the scale pre-estimation model can be obtained through the following steps (1) to (4):

(1) the server obtains a fourth application feature vector of the online plurality of third target applications.

The third target application may be an online target application or a game application in a game knowledge graph. The steps of the server obtaining the fourth application feature vectors of the plurality of online third target application programs are similar to steps 201 to 203, and are not described herein again.

(2) For each third target application, the server obtains a plurality of third user feature vectors for the third target application for the third user.

For each third target application, the step of the server obtaining the third user feature vectors of the plurality of third users for the third target application is similar to the steps 204 to 205, and is not repeated here.

(3) The server acquires the learning labels of a plurality of third users for the third target application program, and the learning label of each third user for the third target application program is used for indicating whether the terminal of the third user runs the third target application program or not.

The server obtains learning labels of a plurality of third users for the third target application program, the learning label of each third user for the third target application program is used for indicating whether the terminal of the third user runs the third target application program, the third user with the learning label of the third user running the third target application program is used as a positive sample, and the third user with the learning label of the third user not running the third target application program is used as a negative sample.

(4) And the server performs model training according to the fourth application characteristic vector of each third target application program, the third user characteristic vector of each third user and the learning label of each third user to obtain a scale pre-estimation model.

In one possible implementation, the server adjusts the proportion of positive and negative samples according to the learning label of each third user. And forming a training set by the fourth application characteristic vector of each third target application program, the third user characteristic vector of each third user and the learning label of each third user, converting the storage format of data in the training set into a TFRecord format, and calling GPU (Graphics Processing Unit) resources to perform model training to obtain a scale estimation model. The GPU resources may also be online GPU resources.

In another possible implementation manner, the server may further obtain attribute information of each third user; determining a fourth user feature vector of each third user according to the attribute information of each third user; and performing model training according to the fourth application characteristic vector of each third target application program, the third user characteristic vector of each third user, the fourth user characteristic vector and the learning label of each third user to obtain a scale pre-estimation model.

The attribute information of the third user may be attribute information of the third user such as gender, age, and the like. The server determines a fourth user feature vector of each third user according to the attribute information of each third user; and the server adjusts the proportion of the positive and negative samples according to the learning label of each third user. And forming a training set by the fourth application characteristic vector of each third target application program, the third user characteristic vector of each third user, the fourth user characteristic vector and the learning label of each third user, converting the storage format of data in the training set into a TFRecord format, and calling GPU (Graphics processing Unit) resources to perform model training to obtain the scale estimation model.

In the embodiment of the invention, when the version of the third target application program is updated or a new third target application program is added into the training set, the server can quickly update the scale pre-estimation model according to the tag data fed back by the questionnaire system.

All the above-mentioned optional technical solutions can be combined arbitrarily to form the optional embodiments of the present invention, and are not described herein again.

Fig. 9 is a schematic structural diagram of a user scale estimation apparatus according to an embodiment of the present invention. Referring to fig. 9, the apparatus includes:

an obtaining module 901, configured to obtain multiple application tags of a first target application program, where each application tag corresponds to multiple options; for each application label, acquiring a plurality of label data corresponding to the application label, wherein each label data comprises a selected target option in a plurality of options corresponding to the application label;

a generating module 902, configured to generate a first application feature vector of a first target application according to a plurality of application tags and tag data corresponding to each application tag;

the obtaining module 901 is further configured to obtain second application feature vectors of the online multiple second target applications and first user weights of the multiple first users for each second target application;

a determining module 903, configured to determine a first user feature vector of each first user according to the second application feature vector of each second target application and the first user weight of each first user for each second target application;

the pre-estimation module 904 is configured to input the first application feature vector and the first user feature vector of each first user into a scale pre-estimation model to obtain a first user scale of the first target application program.

In a possible implementation manner, the pre-estimation module 904 is further configured to obtain attribute information of each first user; determining a second user characteristic vector of each first user according to the attribute information of each first user; and inputting the first application characteristic vector, the first user characteristic vector of each first user and the second user characteristic vector into a scale pre-estimation model to obtain the first user scale of the first target application program.

In another possible implementation manner, each tag data further includes a user identifier of a second user who selects the target option; the generating module 902 is further configured to, for each second user, determine, according to the user identifier of the second user, a target option selected by the second user from the multiple options corresponding to the multiple application tags, to obtain multiple target options; generating a third application characteristic vector corresponding to the second user according to the plurality of target options; and carrying out weighted summation on the third application characteristic vector corresponding to each second user to obtain the first application characteristic vector.

In another possible implementation manner, the establishing module is configured to establish a questionnaire system for each application tag according to the application tag and a plurality of options corresponding to the application tag;

the receiving module is used for receiving a numerical value of the second user for scoring the application label on the questionnaire system;

the selection module is used for selecting a target option corresponding to the numerical value from a plurality of options corresponding to the application label according to the numerical value;

the storage module is used for forming label data by the application identifier, the application label and the target option of the first target application program and storing the label data into a label database;

the obtaining module 901 is further configured to obtain, according to the application identifier and the application tag of the first target application program, a plurality of tag data corresponding to the application tag from a tag database.

In another possible implementation manner, the obtaining module 901 is further configured to obtain statistical data of the first target application;

a determining module 903, configured to determine a statistical feature vector of the first target application according to the statistical data;

the pre-estimation module 904 is further configured to input the statistical feature vector and the first user scale into a later-stage correction model to obtain a second user scale of the first target application program.

In another possible implementation manner, the obtaining module 901 is further configured to obtain fourth application feature vectors of a plurality of third target application programs that have been online; for each third target application program, acquiring a plurality of third user feature vectors and learning labels of the third users for the third target application program, wherein the learning label of each third user for the third target application program is used for indicating whether the terminal of the third user runs the third target application program or not;

In another possible implementation manner, the model training module is further configured to obtain attribute information of each third user; determining a fourth user feature vector of each third user according to the attribute information of each third user; and performing model training according to the fourth application characteristic vector of each third target application program, the third user characteristic vector of each third user, the fourth user characteristic vector and the learning label of each third user to obtain a scale pre-estimation model.

It should be noted that: in the user scale estimation apparatus provided in the foregoing embodiment, when the user scale is estimated, only the division of the functional modules is illustrated, and in practical applications, the function distribution may be completed by different functional modules according to needs, that is, the internal structure of the server is divided into different functional modules, so as to complete all or part of the functions described above. In addition, the user scale estimation apparatus provided in the above embodiment and the user scale estimation method embodiment belong to the same concept, and specific implementation processes thereof are detailed in the method embodiment and are not described herein again.

Fig. 10 shows a block diagram of a terminal 1000 according to an exemplary embodiment of the present invention. The terminal 1000 can be: a smart phone, a tablet computer, an MP3 player (Moving Picture Experts Group Audio layer iii, motion video Experts compression standard Audio layer 3), an MP4 player (Moving Picture Experts Group Audio layer IV, motion video Experts compression standard Audio layer 4), a notebook computer, or a desktop computer. Terminal 1000 can also be referred to as user equipment, portable terminal, laptop terminal, desktop terminal, or the like by other names.

In general, terminal 1000 can include: a processor 1001 and a memory 1002.

Processor 1001 may include one or more processing cores, such as a 4-core processor, an 8-core processor, and so forth. The processor 1001 may be implemented in at least one hardware form of a DSP (Digital Signal Processing), an FPGA (Field-Programmable Gate Array), and a PLA (Programmable Logic Array). The processor 1001 may also include a main processor and a coprocessor, where the main processor is a processor for processing data in an awake state, and is also referred to as a Central Processing Unit (CPU); a coprocessor is a low power processor for processing data in a standby state. In some embodiments, the processor 1001 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content required to be displayed on the display screen. In some embodiments, the processor 1001 may further include an AI (Artificial Intelligence) processor for processing a computing operation related to machine learning.

Memory 1002 may include one or more computer-readable storage media, which may be non-transitory. The memory 1002 may also include high-speed random access memory, as well as non-volatile memory, such as one or more magnetic disk storage devices, flash memory storage devices. In some embodiments, a non-transitory computer readable storage medium in memory 1002 is used to store at least one instruction for execution by processor 1001 to implement the user scale prediction method provided by the method embodiments herein.

In some embodiments, terminal 1000 can also optionally include: a peripheral interface 1003 and at least one peripheral. The processor 1001, memory 1002 and peripheral interface 1003 may be connected by a bus or signal line. Various peripheral devices may be connected to peripheral interface 1003 via a bus, signal line, or circuit board. Specifically, the peripheral device includes: at least one of radio frequency circuitry 1004, touch screen display 1005, camera assembly 1006, audio circuitry 1007, positioning assembly 1008, and power supply 1009.

The peripheral interface 1003 may be used to connect at least one peripheral related to I/O (Input/Output) to the processor 1001 and the memory 1002. In some embodiments, processor 1001, memory 1002, and peripheral interface 1003 are integrated on the same chip or circuit board; in some other embodiments, any one or two of the processor 1001, the memory 1002, and the peripheral interface 1003 may be implemented on separate chips or circuit boards, which are not limited by this embodiment.

The Radio Frequency circuit 1004 is used for receiving and transmitting RF (Radio Frequency) signals, also called electromagnetic signals. The radio frequency circuitry 1004 communicates with communication networks and other communication devices via electromagnetic signals. The radio frequency circuit 1004 converts an electrical signal into an electromagnetic signal to transmit, or converts a received electromagnetic signal into an electrical signal. Optionally, the radio frequency circuit 1004 comprises: an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, a subscriber identity module card, and so forth. The radio frequency circuit 1004 may communicate with other terminals via at least one wireless communication protocol. The wireless communication protocols include, but are not limited to: metropolitan area networks, various generation mobile communication networks (2G, 3G, 4G, and 5G), Wireless local area networks, and/or WiFi (Wireless Fidelity) networks. In some embodiments, the rf circuit 1004 may further include NFC (Near Field Communication) related circuits, which are not limited in this application.

The display screen 1005 is used to display a UI (User Interface). The UI may include graphics, text, icons, video, and any combination thereof. When the display screen 1005 is a touch display screen, the display screen 1005 also has the ability to capture touch signals on or over the surface of the display screen 1005. The touch signal may be input to the processor 1001 as a control signal for processing. At this point, the display screen 1005 may also be used to provide virtual buttons and/or a virtual keyboard, also referred to as soft buttons and/or a soft keyboard. In some embodiments, display screen 1005 can be one, providing a front panel of terminal 1000; in other embodiments, display 1005 can be at least two, respectively disposed on different surfaces of terminal 1000 or in a folded design; in still other embodiments, display 1005 can be a flexible display disposed on a curved surface or on a folded surface of terminal 1000. Even more, the display screen 1005 may be arranged in a non-rectangular irregular figure, i.e., a shaped screen. The Display screen 1005 may be made of LCD (Liquid Crystal Display), OLED (Organic Light-Emitting Diode), and the like.

The camera assembly 1006 is used to capture images or video. Optionally, the camera assembly 1006 includes a front camera and a rear camera. Generally, a front camera is disposed at a front panel of the terminal, and a rear camera is disposed at a rear surface of the terminal. In some embodiments, the number of the rear cameras is at least two, and each rear camera is any one of a main camera, a depth-of-field camera, a wide-angle camera and a telephoto camera, so that the main camera and the depth-of-field camera are fused to realize a background blurring function, and the main camera and the wide-angle camera are fused to realize panoramic shooting and VR (Virtual Reality) shooting functions or other fusion shooting functions. In some embodiments, camera assembly 1006 may also include a flash. The flash lamp can be a monochrome temperature flash lamp or a bicolor temperature flash lamp. The double-color-temperature flash lamp is a combination of a warm-light flash lamp and a cold-light flash lamp, and can be used for light compensation at different color temperatures.

The audio circuit 1007 may include a microphone and a speaker. The microphone is used for collecting sound waves of a user and the environment, converting the sound waves into electric signals, and inputting the electric signals to the processor 1001 for processing or inputting the electric signals to the radio frequency circuit 1004 for realizing voice communication. For stereo sound collection or noise reduction purposes, multiple microphones can be provided, each at a different location of terminal 1000. The microphone may also be an array microphone or an omni-directional pick-up microphone. The speaker is used to convert electrical signals from the processor 1001 or the radio frequency circuit 1004 into sound waves. The loudspeaker can be a traditional film loudspeaker or a piezoelectric ceramic loudspeaker. When the speaker is a piezoelectric ceramic speaker, the speaker can be used for purposes such as converting an electric signal into a sound wave audible to a human being, or converting an electric signal into a sound wave inaudible to a human being to measure a distance. In some embodiments, the audio circuit 1007 may also include a headphone jack.

A location component 1008 is employed to locate a current geographic location of terminal 1000 for navigation or LBS (location based Service). The positioning component 1008 may be a positioning component based on the GPS (global positioning System) in the united states, the beidou System in china, the graves System in russia, or the galileo System in the european union.

Power supply 1009 is used to supply power to various components in terminal 1000. The power source 1009 may be alternating current, direct current, disposable batteries, or rechargeable batteries. When the power source 1009 includes a rechargeable battery, the rechargeable battery may support wired charging or wireless charging. The rechargeable battery may also be used to support fast charge technology.

In some embodiments, terminal 1000 can also include one or more sensors 1010. The one or more sensors 1010 include, but are not limited to: acceleration sensor 1011, gyro sensor 1012, pressure sensor 1013, fingerprint sensor 1014, optical sensor 1015, and proximity sensor 1016.

Acceleration sensor 1011 can detect acceleration magnitudes on three coordinate axes of a coordinate system established with terminal 1000. For example, the acceleration sensor 1011 may be used to detect components of the gravitational acceleration in three coordinate axes. The processor 1001 may control the touch display screen 1005 to display a user interface in a landscape view or a portrait view according to the gravitational acceleration signal collected by the acceleration sensor 1011. The acceleration sensor 1011 may also be used for acquisition of motion data of a game or a user.

The gyro sensor 1012 may detect a body direction and a rotation angle of the terminal 1000, and the gyro sensor 1012 and the acceleration sensor 1011 may cooperate to acquire a 3D motion of the user on the terminal 1000. From the data collected by the gyro sensor 1012, the processor 1001 may implement the following functions: motion sensing (such as changing the UI according to a user's tilting operation), image stabilization at the time of photographing, game control, and inertial navigation.

Pressure sensor 1013 may be disposed on a side frame of terminal 1000 and/or on a lower layer of touch display 1005. When pressure sensor 1013 is disposed on a side frame of terminal 1000, a user's grip signal on terminal 1000 can be detected, and processor 1001 performs left-right hand recognition or shortcut operation according to the grip signal collected by pressure sensor 1013. When the pressure sensor 1013 is disposed at a lower layer of the touch display screen 1005, the processor 1001 controls the operability control on the UI interface according to the pressure operation of the user on the touch display screen 1005. The operability control comprises at least one of a button control, a scroll bar control, an icon control and a menu control.

The fingerprint sensor 1014 is used to collect a fingerprint of the user, and the processor 1001 identifies the user according to the fingerprint collected by the fingerprint sensor 1014, or the fingerprint sensor 1014 identifies the user according to the collected fingerprint. Upon identifying that the user's identity is a trusted identity, the processor 1001 authorizes the user to perform relevant sensitive operations including unlocking a screen, viewing encrypted information, downloading software, paying, and changing settings, etc. Fingerprint sensor 1014 can be disposed on the front, back, or side of terminal 1000. When a physical key or vendor Logo is provided on terminal 1000, fingerprint sensor 1014 can be integrated with the physical key or vendor Logo.

The optical sensor 1015 is used to collect the ambient light intensity. In one embodiment, the processor 1001 may control the display brightness of the touch display screen 1005 according to the intensity of the ambient light collected by the optical sensor 1015. Specifically, when the ambient light intensity is high, the display brightness of the touch display screen 1005 is increased; when the ambient light intensity is low, the display brightness of the touch display screen 1005 is turned down. In another embodiment, the processor 1001 may also dynamically adjust the shooting parameters of the camera assembly 1006 according to the intensity of the ambient light collected by the optical sensor 1015.

Proximity sensor 1016, also known as a distance sensor, is typically disposed on a front panel of terminal 1000. Proximity sensor 1016 is used to gather the distance between the user and the front face of terminal 1000. In one embodiment, when proximity sensor 1016 detects that the distance between the user and the front surface of terminal 1000 gradually decreases, processor 1001 controls touch display 1005 to switch from a bright screen state to a dark screen state; when proximity sensor 1016 detects that the distance between the user and the front of terminal 1000 is gradually increased, touch display screen 1005 is controlled by processor 1001 to switch from a breath-screen state to a bright-screen state.

Those skilled in the art will appreciate that the configuration shown in FIG. 10 is not intended to be limiting and that terminal 1000 can include more or fewer components than shown, or some components can be combined, or a different arrangement of components can be employed.

Fig. 11 is a schematic structural diagram of a server according to an embodiment of the present invention, where the server 1100 may generate a relatively large difference due to a difference in configuration or performance, and may include one or more processors (CPUs) 1101 and one or more memories 1102, where the memory 1102 stores at least one instruction, and the at least one instruction is loaded and executed by the processors 1101 to implement the methods provided by the above method embodiments. Of course, the server may also have components such as a wired or wireless network interface, a keyboard, and an input/output interface, so as to perform input/output, and the server may also include other components for implementing the functions of the device, which are not described herein again.

An embodiment of the present invention further provides a computer-readable storage medium, where at least one instruction, at least one program, a code set, or an instruction set is stored in the computer-readable storage medium, and the instruction, the program, the code set, or the instruction set is loaded by a processor and has an operation in the user scale prediction method for implementing the foregoing embodiment.

It will be understood by those skilled in the art that all or part of the steps of implementing the above embodiments may be implemented by hardware, or may be implemented by a program instructing relevant hardware, where the program may be stored in a storage medium, and the storage medium may be a read-only memory, a magnetic disk, an optical disk, or the like.

The above description is only a preferred embodiment of the present invention, and should not be taken as limiting the invention, and any modifications, equivalents, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for estimating a user size, the method comprising:

2. The method of claim 1, wherein the inputting the first application feature vector and the first user feature vector of each first user into a size prediction model to obtain the first user size of the first target application comprises:

acquiring attribute information of each first user;

3. The method of claim 1, wherein each tag data further includes a user identification of the second user selecting the target option;

4. The method according to claim 1, wherein before the obtaining of the plurality of tag data corresponding to the application tag, the method further comprises:

5. The method of claim 1, wherein after inputting the first application feature vector and the first user feature vector of each first user into a size prediction model to obtain the first user size of the first target application, the method further comprises:

acquiring statistical data of the first target application program;

6. The method of claim 1, wherein before entering the first application feature vector and the first user feature vector of each first user into a size prediction model to obtain the first user size of the first target application, the method further comprises:

7. The method according to claim 6, wherein the model training according to the fourth application feature vector of each third target application, the third user feature vector of each third user and the learning label of each third user to obtain the scale prediction model comprises:

acquiring attribute information of each third user;

8. A user scale prediction apparatus, comprising:

9. A server, characterized in that the server comprises:

a processor and a memory, the memory having stored therein at least one instruction, at least one program, a set of codes, or a set of instructions, the instruction, the program, the set of codes, or the set of instructions being loaded and executed by the processor to implement the operations in the user size prediction method of any of claims 1 to 7.

10. A computer-readable storage medium having stored therein at least one instruction, at least one program, a set of codes, or a set of instructions, which is loaded and executed by a processor to perform the operations performed in the user scale prediction method of any one of claims 1 to 7.